The aim of this in vitro study was to evaluate the effect of the combinations of two different intraoral scanners (IOS), two milling machines, and two restorative materials on the marginal/internal fit and fatigue behavior of endocrowns produced by CAD-CAM. Eight groups (n= 10) were considered through the combination of TRIOS 3 (TR) or Primescan (PS) IOS; 4-axes (CR; CEREC MC XL) or 5-axes (PM; PrograMill PM7) milling machines; and lithium disilicate (LD; IPS e.max CAD) or resin composite (RC; Tetric CAD) restorative materials. Specific surface treatments were applied to each material, and the bonding to its corresponding Endocrown-shaped fiberglass-reinforced epoxy resin preparations was performed (Variolink Esthetic DC). Computed microtomography (μCT) was performed to assess the marginal/internal fit, as well as a mechanical fatigue test (20 Hz, initial load = 100 N/5000 cycles; step-size = 50 N/10,000 cycles until a threshold of 1500 N, then, the step-size was increased if needed to 100 N/10,000 cycles until failure or a threshold of 2800 N) to evaluate the restorations long-term behavior. Complementary analysis of the fracture features and surface topography in scanning electron microscopy was performed. Three-way ANOVA and Kaplan-Meier test (α = 0.05) were performed for marginal/internal fit, and fatigue behavior data, respectively. PS scanner, CR milling machine, and RC endocrowns resulted in a better marginal fit compared to their counterparts. Still, the PM machine resulted in a better pulpal space fit compared to the CR milling machine. Regardless of the scanner and milling machine, RC endocrowns exhibited superior fatigue behavior than LD ones. LD endocrowns presented margin chipping regardless of the milling machine used. Despite minor differences in terms of fit, the 'IOS' and 'milling machine' factors did not impair the fatigue behavior of endocrowns. Resin-composite restorations resulted in a higher survival rate compared to glass-ceramic ones, independently of the digital devices used in the workflow.